The blades are the core components of wind turbines

and their structural damage will not only reduce the efficiency of wind power generation

but also may threaten the overall safety of wind turbines. To address this problem

a Bootstrap-based Hotelling T2 control chart is adopted for condition monitoring of blades. Firstly

the vibration signals of blades are preprocessed to extract fault-related features. And then

the Bootstrap method is used to design the control chart to ensure that it can accurately reflect the blade vibration patterns under normal condition. Finally

the trained control chart is utilized to monitor the state of blades. Compared with classification models and classical control charts

the proposed method can reduce the requirements of data by using only the features of normal condition for learning and determining the threshold of control chart based on Bootstrap method. Practical application shows that the Bootstrap-based Hotelling T2 control chart can effectively detect the state changes of blades.